A beloved convenience store chain on the East coast brought Leverage Lab in to organize their abundant data and drive personalization across their entire user experience. From the gas pump to paid media, right down to the moment of truth at the point-of-sale kiosk they are using weather, geo, transactional, and time-of-day data to make each interaction as personalized as it can be.
As an early proof of concept for this ambitious strategy, we ran a simple A/B test that demonstrated the benefits of unlocking the potential of their data with machine learning. Two audiences were created of users who joined a loyalty reward program. One of the audiences was built from Machine learning modeling to predict those users most likely to convert on an offering.
Both audiences were sent a promotional email with a call to action that said “A Bonus Reward is waiting for you! It’s our special way of saying thanks for being a valued customer and Rewards member. To get your Bonus Reward, use the XXX App or visit XXXRewards.com and log in to your account. Enjoy!” The reward was a free coffee on their next visit.
In other client work, we regularly see a 150% boost in return on ad spend (ROAS) when using machine learning to build lookalike audiences. But in the case of our convenience store client, we used predictive machine learning models to create a segment of current users that were more likely than the mean to convert on an offering. The result saw the machine learning segment generate 2x more conversions than the control audience. No other variables were at play.
Based on these great results, we can’t wait to amp up our client’s personalization by layering on weather, geo, transactional, and time of day data at every user touchpoint.
Click here to download Leverage Lab’s user journey benchmarks: >>> DOWNLOAD >>>